Validations of Blood Pressure Measuring Devices Using Recognized Protocols

Preventing, diagnosing, and controlling high blood pressure is a global health priority. The self-measurement of blood pressure is therefore fundamental and should be done with devices validated by recognized protocols, although most are not. The most widely used and current protocols are the 2010 European Society of Hypertension (ESH) revision and the 2018 Association for the Advancement of Medical Instrumentation (AAMI)/ ESH/ the International Organization for Standardization (ISO) universal standard, respectively. The aim of this study was to find out which blood pressure measuring devices have been adequately validated by the above protocols. A narrative review of blood pressure device validations was conducted by searching the PubMed database. From 52 records identified, 37 studies were included. Most validations follow the 2010 revision and only six follow the 2018 protocol, which is more demanding. Almost all validated sphygmomanometers are automated oscillometric sphygmomanometers in the general population. Wrist devices and devices combining new technologies are also validated, as well as in specific populations, such as the obese, pregnant women, or children. There is sufficient evidence to confirm that the universal AAMI/ ESH/ISO standard is considered the protocol of the century. However, it is necessary to increase the number of validations following it and, above all, validations of the new technologies that are invading the current market.


Introduction
After obesity, arterial hypertension (HBP) is the second most common cause of cardiovascular disease (CVD) [1]. If we add that CVD is the leading cause of morbidity and mortality [2,3], we can conclude that AHT carries a high risk of CV morbidity and mortality [4][5][6][7].
Therefore, one of the main objectives of health systems is to identify people with AHT and ensure that they have good control of their blood pressure (BP), since the higher the blood pressure levels, the greater the risk and morbidity and mortality from CV events [8][9][10]. Thus, preventing, diagnosing, treating, and controlling hypertension is a global health priority [4,5,7,[9][10][11][12][13][14].
The detection and diagnosis of hypertension must be done by measuring BP. BP measurement is considered as one of the most frequently performed procedures at the clinical level, in primary or specialized care [15].
The first time BP was recorded in a consultation was in 1896, and since the end of the last century, out-of-office measurement has been established in two versions: ambulatory BP monitoring (ABPM) and home blood pressure monitoring (HBPM). The major advantage of these two methods is that by allowing BP to be measured on multiple occasions outside the

Search Strategy
The database search was carried out during November-January 2021. Pubmed was the database used for this process.
The advanced search strategy was as follows, combining the terms with the Boolean and grouping operators that follow: (European Society of Hypertension [Title]) OR AAMI/ESH/ISO[Title]) AND validation [Title].
As for the search filters, only one of the publication dates is applied: Last 5 years.
The search strategy has been filtered by title because, by regulation, all validations following the "the revised 2010 European Society of Hypertension international protocol" and "the 2018 AAMI/ESH/ISO universal standard" should be named in a standardized way and, consequently, the titles of these publications should also be named in the same way.

Selection Criteria
Inclusion criteria comprised blood pressure device validation studies that followed the 2010 European Society of Hypertension international protocol review and/or the 2018 AAMI/ESH/ISO universal standard, conducted within the last 5 years.
Exclusion criteria included studies with a publication date prior to 2017, and those that were not device validations or did not follow the revised 2010 European Society of Hypertension international protocol and/or the 2018 AAMI/ESH/ISO universal standard.

Data Extraction
The data obtained were divided according to the validation protocol used. The common methodology governing the validation conditions was recorded, such as sample size, blood pressure range, and other variables, in accordance with the protocol used.
Study characteristics were also recorded, such as citation year and place of validation, types of devices used, population characteristics and origin, and validation conclusions.
Additional information was provided where necessary.

Flow Diagram
From 52 records identified through the searching process, 15 were removed due to exclusion, and finally, 37 studies were included in narrative synthesis ( Figure 1).

Data Extraction
The data obtained were divided according to the validation protocol used. The common methodology governing the validation conditions was recorded, such as sample size, blood pressure range, and other variables, in accordance with the protocol used.
Study characteristics were also recorded, such as citation year and place of validation, types of devices used, population characteristics and origin, and validation conclusions.
Additional information was provided where necessary.

Flow Diagram
From 52 records identified through the searching process, 15 were removed due to exclusion, and finally, 37 studies were included in narrative synthesis ( Figure 1). 15 studies were excluded for the following reasons:  7 articles discussed the validation standard itself and/or revisions of the validation standard  2 studies commented with the proper use of validation protocols  4 publications were corrections of previously included articles.  1 paper was about validation but only followed the ESH 2002 protocol.  1 study did not analyze validity but reproducibility.

Characteristics of the Validation Protocols
The studies of the last five years on blood pressure measuring devices mainly apply two validation protocols for BP devices, which are the review of the international ESH 15 studies were excluded for the following reasons: − 7 articles discussed the validation standard itself and/or revisions of the validation standard − 2 studies commented with the proper use of validation protocols − 4 publications were corrections of previously included articles. − 1 paper was about validation but only followed the ESH 2002 protocol. − 1 study did not analyze validity but reproducibility.

Characteristics of the Validation Protocols
The studies of the last five years on blood pressure measuring devices mainly apply two validation protocols for BP devices, which are the review of the international ESH protocol [27] and the international universal standard AAMI/ESH/ISO 81160-2:2018 [29], the latter being the most recent.
Thus, 31 articles following the ESH review [27] and six based on the AAMI/ESH/ISO universal standard [29] are found in the literature. The methodology of these studies has common characteristics, depending on the validation protocol they follow ( Table 1).
The characteristics of the sample, the sample recruitment criteria, the blood pressure measurement method, or the analyses required to validate the device in question will differ. Both protocols use a similar sequential BP measurement procedure by alternating two devices (reference and test). Validated devices are already used as a reference, but the 2010 ESH prefers these to be two mercury sphygmomanometers with stethoscopes and the AAMI/ESH/ISO standard indicates that the reference can be as stated above, but can also be non-mercury sphygmomanometers, aneroid manometers, or other types. The measurement conditions are similar, where the subject being measured must remain relaxed and calm in a certain position. Also, the human validation equipment is almost identical, and it has a similar tolerable BP measurement error. In both protocols, Bland-Altman charts are required.
On the other hand, they vary substantially in the range of BP recruitment, sample characteristics and size, and validation criteria.
For a device to pass the ESH [27], it must pass two phases. To pass the first phase, two conditions must be met: a minimum of 65, 81, and 93 comparisons falling within 5, 10, and 15 mm/Hg, respectively; a minimum of two of the following three requirements: 73, 87, and 96 differences must be within the category of 5, 10, and 15 mm/Hg, respectively. In the second phase, a minimum of 24 subjects are required to have two of their three differences in the 5 mm/Hg category, and a maximum of three individuals with the three differences greater than 5 mm/Hg is allowed.
To pass the universal standard [29], the device must also pass two phases. In the first phase, the mean BP difference must be 5 mm/hg or less, and its standard deviations 8 mm/hg or less for SBP and DBP. In the second phase, the standard deviation of 85 averaged BP differences (test minus reference BP per subject) must be within a threshold defined by the mean test-reference BP difference listed for SBP and DBP.
In terms of specific populations, the standards proposed by the three societies as a whole are more explicit.
Thus, although the ESH [27] protocol improved on previous protocols [24,26,27] by eliminating some validation steps and reducing the sample size, it is simpler to apply than the most recent protocol. The international universal validation protocol (AAMI/ESH/ISO) of 2018 [29] is currently the most complete but more complex than the previous one.

Characteristics of the Studies
All the studies collected have the same quality in terms of study type, as they all follow the same standards and have the same design: prospective cross-sectional observational studies [31].
Having analyzed the common and different aspects in terms of the validation conditions governing the respective protocols (Table 1), we now turn to the rest of the characteristics, which are summarized in Table 2.

Discussion
We found positive results in most studies, exceeding the protocol in question and validating the test devices for their recommended use, as well as in the population analyzed, with the exception of two studies with four test devices, Omron RS6 (hem-6221-E), Microlife WatchBP O3 [33], Yuwell YE680a, and Cofoe KF-65B [67].
The validated devices are different depending on the measurement method, the area of the body where they are applied, their characteristics, or their function: self-measurement, BP monitoring, or professional use, among others. Only one study has validated a device that uses the auscultatory method to measure BP [66]. Typically, oscillometric upper arm sphygmomanometers are validated as the conventional devices for measuring BP and heart rate (HR). These consist of an automatic monitor with an LCD display and a cuff connected by rubber tubes, with some exceptions, which do not have tubes connecting the pump to the cuff; rather, the device itself is embedded in the cuff [42,47,53,55]. They are usually battery operated and some have various cuff sizes.
The cuff used by most authors in their test device is usually the standard or medium size (22-32 cm), except for some who have tested more than one cuff, such as small (18-22 cm) [65], or large (32-42 cm) [65,66].
There are authors who opt for other wrist devices of the same type [33,35,38,44,50,61] and therefore do not have rubber tubes. These are more applicable to obese subjects because they have a larger arm circumference, giving erroneous measurement results if a standard upper cuff is used [33].
Of the above, many are recommended devices for HBPM but some are designed for ABPM [32,36,46,48,53], while there are few validated sphygmomanometers that are used only in a professional manner [58,59,65,66]. According to the ESH protocol, the validation of ABPM devices is performed only in static conditions and does not require testing in ambulatory conditions. Among those used for HBPM, other validated devices differ more from the above. The Inbody BPBIO320 [49] and the Inbody BPBIO750 [62] are for public use as well as being automatic right upper arm oscillometric devices, but in this case, it is a kiosk type with a fixed hole for the user to insert their arm to measure their BP. Other devices have also been validated that measure BP oscillometrically and can be connected wirelessly to electronic devices allowing for better visual recording and monitoring [46,51,63]. Other units have similar behavior but are operated directly from an app connected to the oscillometric cuff via Bluetooth [42,47,55]. To this effect, one study tests an app that measures BP without a cuff through fingerprints [68] but it is difficult to follow the validation protocol with it.
Therefore, of all the validations, only one device without a cuff to measure BP is attempted to be validated [68], and the results are unsuccessful.
It should be noted that the use of new technologies for health promotion and disease prevention is booming and, in particular, BP control using these has gained importance, adding additional advantages to HBPM, which results in a more active subject with better adherence and follow-up [68][69][70][71]. Despite this and its widespread use and downloading by society, only six apps are being validated [42,46,47,51,55,63,68], considering that one of them, Qardioarm, has been validated on several occasions, by different teams and in different populations [42,47,55]. This is not a unique case, since it draws our attention to the fact that the Omron RS6 is tested on obese subjects and is found to be invalid by some authors [33] and valid by others [61]. This may be due to the fact that both studies follow the ESH review and do not have specific criteria for vulnerable populations [27].
These measured characteristics are generally similar as they all meet the recruitment criteria of the protocols [27,29], except for special populations.
Most of these studies are conducted in the general population, which are healthy adults over 25 years of age for one protocol [27], and 12 years of age for another [29]. In all of them, the number of women recruited tends to be higher than that of men.
Some studies look at other specific populations prone to HBP, such as the obese [33,47,61], children [36,65], diabetics [42], pregnant women [48], or subjects with chronic kidney disease [55]. Furthermore, the difference between the recruited and tested subjects is often greater in these specific populations, as more failures encountered tend to emerge. Most of the reasons for exclusion are full BP ranges. The above are necessary because the results obtained in the general population cannot be extrapolated to specific populations, as they do not follow the same recruitment conditions, requiring a larger sample in most cases. Hence, the universal standard already applies explicit criteria for such populations [29] where even stricter BP control is necessary.
As for the validation team, it also follows the rules of the protocols [27,29], reflected in Table 1, so it usually corresponds to two observers and a supervisor, except for others that have additional explanations, such as teams formed by three licensed physicians [32,48,58], three nurses [47,55], four physicians [49], or three medical technologists [50,59]. All teams were trained in BP measurement.
Although the universal standard is considered the protocol of the century [29], only seven articles have been seen to follow it [56,58,59,62,65,66,68] compared to the large number of studies that follow the ESH [27], almost all of them with positive results [72,73], because although they share conditions, it is less strict ( Table 1).
For example, it was recommended to the journals that, from November 2019 onward, they no longer accept articles following the ESH to validate the devices [74] and only accept the most recent protocol [29], so it is considered necessary to increase the monitoring of this type of device, especially their apps, following the same protocol. In addition, a mercury-free sphygmomanometer can be used as the gold standard in this protocol and is used by the ESH.
To conclude this section, it is worth mentioning that one of the main limitations of our study is that we have only used one database, Pubmed. We were guided by a study that analysed optimal database combinations for literature searches and found that "Sixteen percent of the included references (291 articles) were only found in a single database" [75].

Conclusions
Since 2017, 37 devices have been attempted to be validated to measure BP with reputable protocols, with most obtaining positive results. It is necessary to increase the number of studies according to what is considered the protocol of the century, the universal AAMI/ESH/ISO standard (ISO 81160-2:2018), and especially using new technologies, as few validations have been found on them.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/jpm13010009/s1. Scale S1: Scale for the Assessment of Narrative Review Articles (SANRA). Disclaimer/Publisher's Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.